Author Topic: Antanas's wip's tests and other fails :)  (Read 3332 times)

2014-01-03, 02:21:52

antanas

  • Active Users
  • **
  • Posts: 269
  • Hmm ...
    • View Profile
 Hello everyone! Well I'm new to corona, but here is my current interior design project viz I've done/doing using corona for the first or almost the first time, the design itself is still in active development stage but sadly, now it's more drawing hvac/plumbing and scheduling time (oh how I hate those) so currently I'm unable to make those renders better/more realistic. I know there are plenty of mistakes in design and in viz in particular (damn that constant lack of time) and I wouldn't be posting such shamefull renderings from worked on project, If not for my urgent craving :) for daily builds, which got some really necessary features which I hope will make using corona renderer much easier for me.
So here they are, I hope I didn't screw up too much :) and also hope I will have some time to make em better later.
   

2014-01-03, 11:50:03
Reply #1

Ludvik Koutny

  • VIP
  • Active Users
  • ***
  • Posts: 2562
  • Just another user
    • View Profile
    • My Portfolio
If they had really been shameful, then you would never be posting them there. They look nice...  and only shame on them is the noise, which is not your fault, but fault of corona for still not being fast enough :D

2014-01-03, 12:08:35
Reply #2

antanas

  • Active Users
  • **
  • Posts: 269
  • Hmm ...
    • View Profile
If they had really been shameful, then you would never be posting them there. They look nice...  and only shame on them is the noise, which is not your fault, but fault of corona for still not being fast enough :D

Well thanks, yeah you are right about everything except for corona's speed - it's quite fast, and it is really my fault that renders are noisy - most of them rendered under 1h and some even about under 30 min on a single pc - that's not too far away from say v-rays speed and at least more than few times faster than Maxwell render (to which it gives most similar results in terms of lighting quality) - the only thing which is imho really frustrating in terms of speed is the lack of distributed singleframe rendering capabilities . I did some comparison renders using v-ray 2.4, corona and maxwell 3.0 and will post them a bit later 
« Last Edit: 2014-01-03, 12:12:35 by antanas »

2014-01-03, 12:58:06
Reply #3

antanas

  • Active Users
  • **
  • Posts: 269
  • Hmm ...
    • View Profile
Here are those renders comparison tests , I know conditions are not perfectly the same but nevertheless here they are and I wouldn't under any circumstances say that corona is too slow especially considered that I forgot to use portals in its test :)
Tried to match colors and brightness levels on all of them to be the same - well got get them to be pretty close.

2014-01-03, 14:00:09
Reply #4

Ludvik Koutny

  • VIP
  • Active Users
  • ***
  • Posts: 2562
  • Just another user
    • View Profile
    • My Portfolio
Oh...  so you just did not let them cook long enough. Most of the people have noisy renders because they wait for hours and then give up as the noise does not go away. While sometimes it is problem of a wrong setup, there are still cases when corona simply does not handle given scenario efficiently.

That being said, i have always found some way to make corona perform well :)

2014-01-03, 15:08:05
Reply #5

antanas

  • Active Users
  • **
  • Posts: 269
  • Hmm ...
    • View Profile
That being said, i have always found some way to make corona perform well :)

 - that's in my opinion is that separates good and versatile render engine from a toy , v-ray has such capability, corona has it too, even mental ray has it but for example maxwell, arion, and many others do not, I mean the ability to get good and fast results using different methods :)

2014-07-28, 15:29:49
Reply #6

antanas

  • Active Users
  • **
  • Posts: 269
  • Hmm ...
    • View Profile
Here goes another renders, hopefully not so pitiful as the ones before and even more hopefully not even fails at all, yet I still think something is amiss but at the moment I just can't see what exactly.

2014-07-28, 15:35:34
Reply #7

antanas

  • Active Users
  • **
  • Posts: 269
  • Hmm ...
    • View Profile
As I now see forum image engine (or whatever it is called) "chewed" up those images quite considerably so here goes another try:

tv closeup cam 1 by sprindzhiukas, on Flickr

and

tv closeup cam 2 by sprindzhiukas, on Flickr

2014-07-28, 16:18:57
Reply #8

antanas

  • Active Users
  • **
  • Posts: 269
  • Hmm ...
    • View Profile
Ah, yes forgot to mention that's just a part of my current interior design project, well at least the part rendered to sufficient amount of passes to show it to someone else beside client and the other parts were done as panoramas and imho are tolerable only to the interested party, as decent looking panoramas are a real pain to render at sufficient quality, yet if someone want's to see em' I'll post those too though expect no miracles there. :)
« Last Edit: 2014-07-28, 16:24:44 by antanas »

2014-08-24, 13:49:00
Reply #9

antanas

  • Active Users
  • **
  • Posts: 269
  • Hmm ...
    • View Profile
Hello again, here I've done some more tests regarding the difference in specular/reflections using solid or thin glass and even solid glass plus thin material in windows using hdri for environment lighting, well that will probably apply not for just hdris but for corona sky too though that need to be tested.
thin glass + thin in glass material:
thin glass plus thin glass material by sprindzhiukas, on Flickr
solid glass + solid glass material:
solid glass plus real glass material by sprindzhiukas, on Flickr
and solid glass  + thin in glass material:
solid glass plus thin glass material by sprindzhiukas, on Flickr

This time totally identical scene settings were used opposed to the first test which results I posted in this thread https://corona-renderer.com/forum/index.php/topic,4975.15.html still the new test's result confirm at least a part of the original one's result and I think the main part - thank god I'm not so stupid as I began to believe :). Though I was a bit too hasty to presume there was some difference in lighting quantity as I just simply forgot that at that time when I did that first test (and it was around two months or so ago) I was using reflection override featuring same hdri (as the one in the environment slot) with tweaked gamma and multiplier setting to compensate for this reflection dimming effect and when I discovered it I just removed that override and tweaked that hdri in environment for new situation hence the difference in lighting appeared + those images were newer intended as a test which I would post/share and some intermediate result was just not saved - sorry bout that, disinformation wasn't intentional :)
« Last Edit: 2014-08-24, 13:57:44 by antanas »